Guide
New brand, fresh ad account, zero pixel data, no followers to poll. The usual advice — "just test it live" — assumes an audience you don't have. Here's how cold-start brands get a real read on creative anyway.
Every "how to test your ads" guide quietly assumes you're already running. They tell you to read your hook rate, watch your CPA, and kill the losers. But on a brand-new account the algorithm has no conversion history to optimize against, your audience is whoever the platform guesses, and your first few hundred dollars buy data so noisy it's almost useless. You can't A/B test your way to a winner when every variant is being shown to a different random slice of strangers. That's the cold-start problem: the methods that work for established accounts need exactly the thing you don't have yet — an audience and a baseline.
There's a second thing you can test that exists on day one: the creative itself. A scroll-stopping opening, a clear middle, and a CTA that lands are properties of the ad, not of your audience size. Pre-testing reads those properties directly. PreTestAds predicts attention and engagement from your ad's audio-visual features and scores it as a percentile against a benchmark of 76 top-performing ads from the TikTok Creative Center — no audience, no pixel, no spend required to get the first read. It won't tell you who will buy, but it will tell you whether your creative is likely to hold attention long enough to matter, which is the question that comes first anyway. See the full approach in ad pre-testing and the deeper read in ad creative analysis.
Make three to five distinct creative concepts — not three edits of the same idea, genuinely different hooks and angles. Score each one for predicted attention before you spend a cent. Look at Hook Strength first; a concept that loses viewers in the first three seconds won't recover with budget. Then read the full curve: where attention drops, the peak moment, and whether engagement survives into the CTA window. Launch only the strongest one or two, put your limited first budget behind them, and let those real impressions start building the conversion data the platform needs to actually optimize. Now you've turned a cold start into a warm one without lighting money on fire to do it. If budget is the real constraint, pair this with the small-budget testing framework and the pre-launch ad checklist.
Be honest with yourself about what you're buying. Predicted attention is a screen, not a sales forecast — it flags creative that's likely to be ignored, so you stop funding it, but it doesn't promise conversions or ROAS. Those still come from your offer, your landing page, your pricing, and real audience response. The right mental model is a goalkeeper: pre-testing blocks the obviously weak shots before they cost you, and live A/B testing decides the winner once you finally have an audience to test against. For a new account, that order matters: screen first so your scarce early spend goes behind creative that at least clears the attention bar.
Upload your first concept and see predicted attention in minutes — first analysis free, no credit card.
Score Your First Ad Free